Restricted Transfer learning for Text Categorization
Rajhans Samdani, Gideon Mann · 2013
In practice, machine learning systems deal with multiple datasets over time. When the feature spaces between these datasets overlap, it is possible to transfer infor-mation from one task to another. Typically in transfer learning, all labeled data from a source task is saved to be applied to a new target task thereby raising con-cerns of privacy, memory and scaling. To ameliorate such concerns, we present a semi-supervised algorithm for text categorization that transfers information across tasks without storing the data of the souce task. In particular, our technique learns a sparse low-dimensional projection from unlabeled and the source task data. In particular, our technique learns low-dimensional sparse word clusters-based fea-tures from the source task data and a massive amount of additional unlabeled data. Our algorithm is efficient, highly parallelizable, and outperforms competi-tive baselines by up to 9 % on several difficult benchmark text categorization tasks. 1